288 HYDROIDiE. Part IV. 



of the last stage. As the pedicel is not elongate, and shows no signs of prep- 

 aration for the development of a fourth medusa, we conclude that three is the 

 highest number on any one axis; we have, at least, never seen more than three. 

 The terminal medusa of Fig. 19 is very nearly ripe, but not quite so, as may be 

 seen by comparing the pear-shaped spermatic particles {Fig. 20, A B), which were 

 taken from it, with those which were naturally discharged from a fully mature 

 animal {Fig. 21, A B). From this period to maturity the spermatic contents have 

 an opaque white color. The mature sj)ermatic particle {Fig. 21, A B) has an 

 elongated, guitar-shaped bodj^, and from its narrower end a tail tapers away, and 

 extends to about eight times its length.' The anterior end suddenly narrows into 

 a rounded prominence. We have searched in vain for female medusiB among this 

 kind of Hydroids. 



No. II. Proles hjdroidea. — The h3'dra of this hydromedusarium (PI. XXVII. 

 Figs. 22-26) can in no way be distinguished from that of No. I., except by its 

 yellow color ; it has the same habitat, size, proportions, and mode of branching, 

 and the structural details do not differ. In the breeding season, however, the 

 differences are very obvious; then we find that the medusse {md) are not only 

 single, but scattered along the stem of the hydra for a considerable distance 

 from the head. As these medusa? were observed in July, the middle of the 

 breeding season of No. I., and the two appeared to be in the same stage of 

 development, they could not be supposed to represent two different states of 

 . the same species ; the less so since the medusa? of both seemed to be males, 

 and the oldest of the second kind were very opaque and yellow, like the whole 

 hydra. Unlike the Eudendrium ramosum figured by VanBeneden and others, the 

 medusfB of our s2Decies do not free themselves, but are developed as simple, saccate, 

 globular bodies, with what appears to be a broad proboscis {Fig. 22, md^). The 

 tendency to form branched bunches, as represented by one of the figures {Fig. 22, 

 a b), indicates a close relation to the type of the true Eudendrium ramosum. 



Recent observations have shown that, notwithstanding the extraordinary differ- 

 ences noticed between the two Hydroids described above, they belong to one and 

 the same species ; the first form being the stock which produces proliferous male 

 medusas only, and the second that which produces single, scattered, female medusae ; 

 there is, besides, a marked difference in the color of the hydroids. This case shows, 

 pei'haps, more full}' than any other, with what perseverance the Hydroids must be 

 studied, in their various stages of growth, before correct results can be reached. 



The buds of the female medusa? are usually scattered irregularly upon the 

 calyx of the hydra ; occasionally they are found at some distance from the head, 

 along the stem. The wall of the bud, which, in the early stages, is of uniform 



